Search Images Maps Play YouTube News Gmail Drive More »
Sign in
Screen reader users: click this link for accessible mode. Accessible mode has the same essential features but works better with your reader.

Patents

  1. Advanced Patent Search
Publication numberUS6993100 B2
Publication typeGrant
Application numberUS 09/735,369
Publication dateJan 31, 2006
Filing dateDec 12, 2000
Priority dateDec 12, 2000
Fee statusPaid
Also published asUS20020168034
Publication number09735369, 735369, US 6993100 B2, US 6993100B2, US-B2-6993100, US6993100 B2, US6993100B2
InventorsGanning Yang, John Walley
Original AssigneeSkyworks Solutions, Inc.
Export CitationBiBTeX, EndNote, RefMan
External Links: USPTO, USPTO Assignment, Espacenet
Burst detector
US 6993100 B2
Abstract
A system for detecting a burst in a wireless communications system. A signal strength indicator indicates the strength of an incoming signal representative of incoming packets. A signal strength change detector detects changes in the signal strength of the incoming signal. Signal strength detection logic determines if a change in signal strength of a predetermined magnitude has occurred. A pattern detector detects patterns of symbols, or symbol estimates, in the incoming signal to determine if a predetermined pattern of symbols is present. Burst detection logic signals detection of a burst if the signal strength detection logic determines that a change in signal strength of predetermined magnitude has occurred, and the pattern detector determines that a predetermined pattern of symbols is present.
Images(13)
Previous page
Next page
Claims(12)
1. A burst detection system for detecting a burst in an incoming signal comprising:
a signal strength change detector for determining strength changes in the incoming signal;
signal strength detection logic for determining if a change in signal strength of a predetermined magnitude has occurred;
a pattern detector for monitoring patterns of symbols in the incoming signal to determine if a predetermined pattern is present;
a burst detector for signaling a detection of a burst if the signal strength change detection logic determines that a signal strength change of predetermined magnitude has occurred and the pattern detector determines that a predetermined pattern of symbols is present;
a signal strength indicator for indicating the strength of the incoming signal;
a short-term signal strength change detector for determining, responsive to the signal strength indicator, short-term changes in signal strength, and a long-term signal strength change indicator for determining, responsive to the signal strength indicator, longterm changes in signal strength;
wherein the signal strength detection logic is configured to determine, responsive to the short-term and long-term signal strength change detectors, if a short-term change in signal strength of a predetermined magnitude has occurred, and a long-term change in signal strength of a predetermined magnitude has occurred;
a signal strength indicator for indicating the strength of the incoming signal, and the short-term and long-term signal strength change detectors respectively determine short-term and long-term changes in signal strength responsive to the indication of signal strength provided by the signal strength indicator;
wherein the short-term signal strength change detector is configured to determine An, a current moving average of M samples of an, the indication of signal strength provided by the signal strength indicator, and Bn, a previous moving average of M samples of an, where M is a non-negative integer.
2. The system of claim 1 wherein the signal strength change detection logic is configured to determine if a short-term change in signal strength of sufficient magnitude has occurred by determining if the ratio of An to Bn exceeds a predetermined threshold.
3. The system of claim 1 wherein the short-term signal strength change detector is configured to determine Cn, a long-term average of an, in accordance with the following expression: Cn=(1−σ)*Cn-1+an, where σ is less than or equal to 1, and indicates the relative weights to be given to Cn-1 and an in the computation of Cn.
4. The system of claim 3 wherein the signal strength change detection logic is configured to determine if a change in signal strength of predetermined magnitude has occurred by determining if the ratio of An to Cn exceeds a predetermined magnitude.
5. A burst detection system for detecting a burst in an incoming signal comprising:
a signal strength change detector for determining strength changes in the incoming signal;
signal strength detection logic for determining if a change in signal strength of a predetermined magnitude has occurred;
a pattern detector for monitoring patterns of symbols in the incoming signal to determine if a predetermined pattern is present;
a burst detector for signaling a detection of a burst if the signal strength change detection logic determines that a signal strength change of predetermined magnitude has occurred and the pattern detector determines that a predetermined pattern of symbols is present;
a signal strength indicator for indicating the strength of the incoming signal;
a short-term signal strength change detector for determining, responsive to the signal strength indicator, short-term changes in signal strength, and a long-term signal strength change indicator for determining, responsive to the signal strength indicator, longterm changes in signal strength;
wherein the signal strength detection logic is configured to determine, responsive to the short-term and long-term signal strength change detectors, if a short-term change in signal strength of a predetermined magnitude has occurred, and a long-term change in signal strength of a predetermined magnitude has occurred;
a symbol detector for detecting symbols, or estimates thereof, in the incoming signal, and the pattern detector monitors the symbols or estimates provided by the symbol detector to determine if a predetermined pattern of symbols is present;
wherein the incoming signal is a quadrature baseband signal, and the symbol detector determines soft estimates δθn of the symbols; and
a symbol spaced differentiator for determining, responsive to the samples δθn from the symbol detector, δδθn =δθn −δθn-L where L is the number of samples/symbol.
6. The system of claim 5 wherein the pattern detector determines if a predetermined pattern of symbols is present responsive to the values δδθn from the symbol spaced differentiator.
7. A method for detecting a burst in an incoming signal comprising:
monitoring short-term signal strength changes in the incoming signal to determine if a short-term change in signal strength of predetermined magnitude has occurred;
monitoring long-term signal strength changes in the incoming signal to determine if a long-term change in signal strength of predetermined magnitude has occurred;
monitoring patterns of symbols in the incoming signal to determine if a predetermined pattern is present;
performing the foregoing three monitoring steps in parallel;
signaling detection of a burst if a short-term signal strength change of predetermined magnitude has occurred, a long-term signal strength change of predetermined magnitude has occurred, and a predetermined pattern of symbols is present;
indicating the strength of the incoming signal, and monitoring short-term and long-term changes in signal strength responsive to the indication of signal strength; and
determining An, a current moving average of M samples of an, the indication of signal strength, and Bn, a previous moving average of M samples of an, where M is a non-negative integer.
8. The method of claim 7 further comprising determining if a short-term change in signal strength of sufficient magnitude has occurred by determining if the ratio of An to Bn exceeds a predetermined threshold.
9. The method of claim 7 further comprising determining Cn, a long-term average of an, in accordance with the following expression: Cn=(1-σ)Cn-1+σan, where a is less than or equal to 1, and indicates the relative weights to be given to Cn-1 and an in the computation of Cn.
10. The method of claim 9 further comprising determining if a change in signal strength of predetermined magnitude has occurred by determining if the ratio of An to Cn exceeds a predetermined magnitude.
11. A method for detecting a burst in an incoming signal comprising:
monitoring short-term signal strength changes in the incoming signal to determine if a short-term change in signal strength of predetermined magnitude has occurred;
monitoring long-term signal strength changes in the incoming signal to determine if a long-term change in signal strength of predetermined magnitude has occurred;
monitoring patterns of symbols in the incoming signal to determine if a predetermined pattern is present;
performing the foregoing three monitoring steps in parallel;
signaling detection of a burst if a short-term signal strength change of predetermined magnitude has occurred, a long-term signal strength change of predetermined magnitude has occurred, and a predetermined pattern of symbols is present;
detecting symbols, or estimates thereof, in the incoming signal, and monitoring the symbols or estimates to determine if a predetermined pattern of symbols is present;
wherein the incoming signal is a quadrature baseband signal, and the method further comprises determining soft estimates δθn of the symbols; and
determining, responsive to the samples δθn, δδθn=δθn−δθn-L, where L is the number of samples/symbol.
12. The method of claim 11 further comprising determining if a predetermined pattern of symbols is present responsive to the values δδn.
Description
BACKGROUND OF THE INVENTION

1. Field of the Invention

This invention relates generally to frequency and/or timing acquisition in a wireless communications system, and, more specifically, reliable burst detection as an antecedent to frequency and/or timing acquisition in a wireless communications system.

2. Related Art

In wireless communications systems, a recipient of a wireless transmission, be it a mobile station or a land station or some other element, must become aware of the timing of and frequency with which the transmission occurred in order to decipher the information in the transmission. Such a process is often referred to as the recipient “acquiring” the frequency or timing of the transmission.

Conventional approaches for timing and/or frequency acquisition include closed loop techniques such as Symbol Timing Recovery (STR) or Automatic Frequency Control (AFC). In these techniques, a transmission is prefaced with a preamble, and the recipient acquires frequency and/or timing by analyzing the preamble.

The problem with these techniques is that the acquisition process can take an inordinate amount of time if there is a large frequency offset between the initial assumed frequency and the actual frequency of the transmission, or there is a large timing error between the initial assumed timing and the actual timing of the transmission. Another problem is that, because data cannot be interpreted until acquisition occurs, there is a danger that data will be received before acquisition occurs, and therefore lost.

The problem can be minimized or corrected by increasing the size of the preamble, but that will degrade system throughput. Moreover, long preambles are also not possible in systems employing protocols such as Bluetooth which are based on standards requiring short preambles. Moreover, because of frequency hopping, in which different packets are transmitted at different frequencies, systems employing protocols such as Bluetooth cannot accommodate long preambles because of the adverse effect that would have on system throughput.

Conventional approaches also include open loop techniques such as that employed in PHS (Personal Handyphone System). An open loop system requires that the recipient be aware of an initial frequency estimation or initial timing estimation. These systems also require that the recipient be aware of the starting point of data transmission. Such a requirement is a significant limitation because it requires that all transmissions be synchronized, which, as a practical matter, may not possible in many systems today.

Some systems, such as TDMA systems, utilize burst detection techniques to detect the timing of an incoming transmission. However, many of these techniques are not reliable in the case of a short preamble as that employed in Bluetooth, which uses only 4 symbols in the preamble.

SUMMARY

The invention provides a burst detection system where one or more power change detectors and one or more pattern detectors are jointly used to detect incoming bursts in the preambles of incoming packets. Detection of an incoming burst then triggers frequency acquisition and demodulation of the bodies of the incoming packets, thereby allowing recovery of the underlying data.

The one or more power change detectors may include a short-term power change detector and a long-term power change detector. A pattern detector may also be included to operate in parallel with the short-term and long-term power detectors. A signal representing the incoming packets may be simultaneously input into the short-term and long-term power detectors, and the pattern detector.

Power change detection logic may also be included to determine if a short-term power change in the incoming signal as detected by the short-term power change detector is of a predetermined magnitude and if a long-term power change in the incoming signal as detected by the long-term power change detector is of a predetermined magnitude. In one example, long-term power detection occurs by monitoring changes in current short-term power versus previous long-term power, but it should be appreciated that other examples are possible.

A pattern detector operating in parallel with the short-term and long-term power change detectors determines if a predetermined pattern of bits is present in the incoming signal. Burst detection logic then signals the detection of a burst if the power change detection logic signals that short and long-term power changes of sufficient magnitude have occurred in the incoming signal, and if the pattern detector detects the predetermined pattern of bits in the incoming signal. Detection of a burst triggers an acquisition and demodulation block to acquire frequency and begin demodulation of the remainder of the packet.

In a Bluetooth implementation, an instantaneous received signal strength indicator (RSSI) block detects the instantaneous RSSI of an incoming quadrature baseband signal. A short-term power detector, which may be implemented as an M-sample boxcar filter, samples the instantaneous RSSI over M samples, and produces a short-term moving average from the M samples. An M-tap delay line may provide the moving average determined M samples ago. A long-term power detector, which may be implemented as an exponential window filter, produces a long-term average of the instantaneous RSSI by producing a weighted average of the instantaneous RSSI and the previous value of the long-term average.

The power change detection logic monitors the ratio between the current and previous short-term moving averages to determine if the ratio exceeds a predetermined threshold. It also monitors the ratio between the current short-term moving average and the long-term weighted average to determine if that ratio exceeds a predetermined threshold. If both conditions are met, the power change detection logic signals this occurrence on an output signal.

A differential phase detector receives as an input the quadrature baseband signal, and produces therefore a differential phase signal δθn. A symbol spaced differentiator receives as an input the differential phase signal δθn, and further differentiates it to eliminate frequency offset. The signal which is produced, δδθn, is given by the following expression: δδθn=δθn−δθn-L, where L, the oversampling rate, in terms of number of samples per symbol. In current Bluetooth implementations, the oversampling rate L is set to 4.

In current Bluetooth implementations, the packet preambles embody a predetermined pattern of 4 symbols, either 1010 or 0101. Consequently, in an environment free from noise, the absolute value of two successive values of δδθn, when determined in the packet preamble, should be 0, and the signs of the two values should be of opposite polarity.

Pattern detection logic detects if two successive values of δδθn, as determined over the packet preambles, is less than a predetermined threshold determined to take account of noise. If this condition is met, and the signs of the two values are of opposite polarity, the pattern detection logic signals the occurrence of these two conditions on an output signal.

A moving window accumulator sums the output of the pattern detection logic over N samples. The number N is the number of samples represented by the packet preamble. In current Bluetooth implementations, the number N is 16, representing 4 samples/symbol over 4 symbols.

A peak detector detects whether the output of the moving window accumulator has reached a peak and exceeds a predetermined threshold, signifying that a predetermined pattern of symbols has been detected in a packet preamble. If so, the peak detector signals the occurrence of this condition on an output signal.

Burst detection logic receives as inputs the output signal of the peak detector, and the output signal of the power change detection logic. If the output of the moving window accumulator has peaked and exceeded a predetermined threshold, as detected by the peak detector, and if both short-term and long-term power changes in the incoming signal have occurred as detected by the power change detection logic, burst detection logic signals the occurrence of a burst.

The burst detection logic signals this condition to the acquisition and demodulation block. This block receives as an input the output δθn of the differential phase detector after passage through a delay compensation buffer. The delay compensation buffer compensates for the delay inherent in the foregoing pattern and power change detection circuitry. Upon detecting a burst, the acquisition and demodulation block acquires frequency and begins demodulating the differential phase input δθn from the differential phase detector. That permits the underlying data in the bodies of the packets to be recovered.

Other systems, methods, features and advantages of the invention will be or will become apparent to one with skill in the art upon examination of the following figures and detailed description. It is intended that all such additional systems, methods, features and advantages be included within this description, be within the scope of the invention, and be protected by the accompanying claims.

BRIEF DESCRIPTION OF THE FIGURES

The invention can be better understood with reference to the following figures. The components in the figures are not necessarily to scale, emphasis instead being placed upon illustrating the principles of the invention. Moreover, in the figures, like reference numerals designate corresponding parts throughout the different views.

FIG. 1 is a block diagram of an embodiment of a burst detection system according to the invention.

FIG. 2 illustrates example packet formats in an example Bluetooth environment.

FIG. 3 is a block diagram of a burst detection system configured for use in a Bluetooth environment.

FIG. 4 is a block diagram of an implementation example of an M-sample boxcar filter for use in a short-term power detector.

FIG. 5 is a block diagram of an M-tap delay line for use in conjunction with the short-term power detector of FIG. 4 to detect short-term power changes.

FIG. 6 is a block diagram of an implementation example of an exponential window filter for use in a long-term power detector.

FIG. 7 is a block diagram of an implementation example of a symbol spaced differentiator for use in a burst detection system configured for a Bluetooth environment.

FIG. 8 is a block diagram of an implementation example of pattern detection logic block configured for a burst detection system in a Bluetooth environment.

FIG. 9 is a block diagram of an implementation example of a moving window accumulation block configured for a burst detection system in a Bluetooth environment.

FIG. 10 is a block diagram of an implementation example of a peak detector configured for a burst detection system in a Bluetooth environment.

FIG. 11 illustrates a representation of an example signal embodying a 4 bit preamble in a Bluetooth environment.

FIG. 12 is a flowchart illustrating an embodiment of a method of operation of the burst detection system according to the invention.

DETAILED DESCRIPTION

The subject invention includes a burst detection system where one or more power change detectors and one or more pattern detectors are jointly used to detect incoming bursts in the preambles of incoming packets. Detection of an incoming burst then triggers frequency and timing acquisition and demodulation of the bodies of the incoming packets, thereby allowing recovery of the underlying data.

A first embodiment of a burst detection system in accordance with the invention is illustrated in FIG. 1. As illustrated, an incoming signal embodying incoming packets is input in parallel to soft-bit detector 104 and power detector 112. Soft-bit detector 104 is configured to detect symbols in the incoming symbol 102 and power detector 112 is configured to detect the power of the incoming signal.

In one implementation, the incoming signal may comprises a quadrature baseband signal having I and Q components, and the soft-bit detector 104 may comprises a differential phase detector which converts the quadrature baseband signal into the theta (θ) domain, and then differences the resulting signal to produce phase differences δθ represented by the incoming signal. The phase differences δθ constitute soft estimates of the underlying symbols, and can be converted into symbols through comparison with a predetermined threshold.

Alternatively, the soft-bit detector 104 may comprise an FM detector which converts the quadrature baseband signal into the theta domain, and then differentiates the resulting signal to produce frequency values. These frequency values again constitute soft estimates of the underlying symbols, and can be converted into symbols through comparison with a predetermined threshold. One of ordinary skill in the art will appreciate, from a reading of this disclosure, that other forms of symbol detection may be possible.

The power detector 112 may comprise in one implementation an instantaneous received signal strength indicator (RSSI) block which detects the instantaneous strength of the incoming signal 102. One of ordinary skill in the art will appreciate from a reading of this disclosure that other forms of power or signal strength detection may be possible.

The output of the power detector in this embodiment is coupled in parallel to a short-term power change detector 114 and a long-term power change detector 116. The short-term power change detector 114 monitors short-term changes in the power or strength of the incoming signal as detected by the power detector 112, and the long-term power change detector 116 monitors long-term changes in the power or strength of the incoming signal as detected by the power detector 112.

Power change detection logic 118 determine if a short-term power change in the incoming signal as detected by the short-term power change detector 114 is of a predetermined magnitude and if a long-term power change in the incoming signal as detected by the long-term power change detector is of a predetermined magnitude. In one implementation, power change detection logic 118 determines whether or not a short-term power change as detected by the short-term power change detector 114 exceeds a predetermined threshold, and whether or not a long-term power change as detected by the long-term power change detector 116 exceeds a predetermined threshold. These thresholds are determined to avoid “false alarm” situations and distinguish over power changes due to the presence of noise. If both conditions are present, the power change detection logic 118 signals the occurrence of this situation on an output signal.

Pattern detector 120 operates in parallel with the short-term and long-term power change detectors 114, 116 to determine if a predetermined pattern of symbols is present in the packet preamble embodied in the incoming signal. The pattern detector 120 receives the symbols or estimates produced by the symbol detector, and monitors the same to determine if the predetermined symbol pattern is present. If the predetermined symbol pattern is detected, pattern detector 120 signals the occurrence of this condition on an output signal.

Burst detection logic monitors the signals output from the power change detection logic 118, and the pattern detector 120. If the signals indicate (a) that the predetermined pattern of symbols has been detected; and (b) both short- and long-term power changes of sufficient magnitude have been detected, then the burst detection logic 122 signals on an output signal the detection of a burst. Detection of a burst triggers an acquisition and demodulation block 108 to acquire frequency and begin demodulation of the remainder of the packet. To ensure proper synchronization between the timing with which the bodies of the packets appear at the acquisition and demodulation block 108 and the detection of a burst through the packet preambles, and compensate for delay through the foregoing burst detection circuitry, a delay element 106 is provided.

In one example application, the system of FIG. 1 is employed in an environment consistent with the Bluetooth Wireless Technology Standard (hereinafter referred to as “Bluetooth”). Bluetooth is a computing and telecommunications industry specification that describes how mobile phones, home and business phones, computers, and PDAs (personal digital assistants) can easily interconnect with each other using a short-range wireless connection. Bluetooth is specifically designed to provide low-cost, robust, high-capacity voice and data networking. It features fast frequency hopping to avoid interference and short data packets to maximum capacity during interference.

In Bluetooth, information is conveyed in the form of packets. The format of a standard Bluetooth packet 200 is illustrated in FIG. 2. The standard packet 200 includes a 72-bit access code 202, a 54-bit header 204 and a payload 206 of variable length ranging from zero to 2,745 bits. The 72-bit access code 202 includes a 4-bit preamble 208, a 64-bit sync word 210, and a 4-bit trailer 212. The 72-bit access code 202 is generally used for synchronization, DC offset compensation and identification, such as between different senders.

FIG. 3 illustrates an embodiment of a burst detector configured for use in a Bluetooth environment. The purpose of the burst detector in this embodiment is to accurately determine the burst starting point. The detector continuously monitors the incoming signal 302 to see if a burst is incoming. If a burst is detected, as will be seen, acquisition and demodulation block 310 is triggered to commence frequency/timing acquisition and data recovery.

In this embodiment, a quadrature baseband signal 302 having I and Q components is input to A/D converter 304. A/D converter 304 converts the signal into the digital baseband. The digital baseband signal is input in parallel to the differential phase detector 306 and the instantaneous received signal strength indicator (RSSI) block 322.

Instantaneous RSSI block 322 detects the instantaneous RSSI of the incoming digital quadrature baseband signal, and outputs a signal an representative of this instantaneous RSSI. In one implementation, this instantaneous RSSI output signal an may be derived from the I and Q components of the baseband signal, and optionally scaled using the scale factor SCALE as determined by RF AGC setting information. The determination of an in this implementation may be represented by the following expression:
a n=max{|I n |, |Q n|}+0.5*min{|I n |, |Q n|}*SCALE.  (1)

Short-term RSSI detector 324 samples the instantaneous RSSI signal an over a relatively small number of samples to reduce noise variance. In one example, short-term RSSI detector 324 may be implemented as an M-sample boxcar filter, which samples the instantaneous RSSI over M samples, and produces an output signal An representing a short-term moving average of the M samples.

An example implementation of such a boxcar filter is illustrated in FIG. 4. The instantaneous RSSI signal an is input to a series combination of M storage elements 400(0), 400(1), . . . , 400(M-1). The storage elements are linked together such that successive samples of an successively pass through the storage elements in succession. The output of each storage element 400(0), 400(1), . . . , 400(M-1) is input to summer 402. Summer 402 outputs the sum of the contents of the storage elements, and divider 404 divides the sum by M, thus producing a moving average An computed over M samples of the input signal an. In one Bluetooth implementation, the parameter M is set to 4, the number of samples in a packet preamble.

Turning back to FIG. 3, in the burst detector, a delay line 328 provides a previous value Bn of the moving average An. In the implementation where short-term RSSI detector 324 is implemented as an M-sample boxcar filter, delay line 328 may be implemented as an M-tap delay line configured to output the moving average from detector 324 determined M samples ago. An example implementation of an M-tap delay line is illustrated in FIG. 5. The current M-sample moving average An from detector 324 is input to a series combination of M storage elements, 500(0), 500(1), 500(2), . . . , 500(M-1). The storage elements are coupled in series so that successive values of An pass successively through the storage elements. The output of the Mth storage element 500(M-1) is Bn, representing An delayed by M samples.

In the burst detector of FIG. 3, long-term power detector 326 provides an output signal Cn representative of a long-term average of the instantaneous RSSI signal an output by instantaneous RSSI block 322. In one example, long-term power detector 326 may be implemented as an exponential window filter, which outputs a weighted average of the instantaneous RSSI signal an and a previous value Cn-1 of the long-term average. The determination of Cn by such a filter may be represented by the following formula:
C n=(1−α)*C n-1 +α*a n  (2)
where an is the input instantaneous RSSI, Cn is the long term RSSI output, and α is the effective time period over which the averaging occurs. In one Bluetooth implementation, the parameter α is set to 1/32.

Power change detection logic 330 monitors the ratio between the current and previous short-term moving averages, An and Bn, to determine if the ratio exceeds a predetermined threshold. It also monitors the ratio between the current short-term moving average An and the long-term weighted average Cn to determine if that ratio exceeds a predetermined threshold. If both conditions are met, the power change detection logic 330 signals this occurrence on an output signal Dn.

In one implementation, the decision logic for the power change detection logic 330 can be represented by the following pseudo-code:
If(An>T1*Bn and An>T2*Cn) then

    • The Power Change Detector output Dn is set to high for a fixed length window.

Else

    • The power change detector output Dn is set to low.
      where T1 and T2 are predetermined thresholds selected based on specific application requirements such as operational SNR, misdetection rate and false alarm rate. In one Bluetooth implementation, T1 and T2 are both set to 2.

Differential phase detector 306 receives as an input the digital quadrature baseband signal, and produces therefrom a differential phase signal δθn. Symbol spaced differentiator 314 receives as an input the differential phase signal δθn, and further differentiates it to eliminate frequency offset. The signal which is produced, δδθn, is given by the following expression: δδθn=δθn−δθn-L, where L, the oversampling rate, in terms of number of samples per symbol. In one Bluetooth implementation, the oversampling rate L is set to 4.

Currently, Bluetooth requires that the packet preambles embody a predetermined pattern of 4 symbols, either 1010 or 0101. Consequently, in an environment free from noise, the absolute value of two successive values of δδθn, when determined in the packet preamble, should be 0, and the signs of the two values should be of opposite polarity.

These conditions can be further understood through consideration of FIG. 11, which illustrates two successive samples of δδθn, A and B, which are spaced by a symbol time period T, in an assumed noise free environment. In this figure, the predetermined pattern which is assumed is 1010. This pattern forms a sine wave, and FIG. 6 illustrates this sine wave after it has been differentiated twice. As can be seen, the result is still a sine wave, with any frequency offset eliminated.

With respect to the two samples A and B, it will be observed that the absolute value of the sum of the amplitudes of the two samples is zero, and their signs are of opposite polarity. It should be noted that these conditions will be present for any two successive samples of δδθn, such as C and D, which are spaced by the symbol time period T.

Due to the presence of noise, however, it may be necessary to slightly relax one or more of these conditions, particularly the condition that the absolute value of two successive symbols sum to zero. Consider, for example, a situation where, due to the presence of noise, the magnitude of the sample A is perturbed to A′=A+ε. Due to this perturbation, the magnitudes of the two samples A and B no longer sum to 0, but to ε. Consequently, to take account of the presence of noise, a relaxation of this condition may be warranted.

Turning back to the burst detector of FIG. 3, pattern detection logic 316, in accordance with the foregoing principles, detects if two successive values of δδθn, as determined over the packet preambles, is less than a predetermined threshold determined to take account of noise. If this condition is met, and the signs of the two values are of opposite polarity, the pattern detection logic signals the occurrence of these two conditions on an output signal dn. In one implementation, the determination of the output dn by pattern detection logic 316 can be represented by the following pseudo-code:
If (|δδθn+δδθn-L |<T 3 and δδθn*δδθn-L<0) then
dn=high
Else
dn=low
where δδθn is the current output of differentiator 314, and δδθn-L is the previous output of differentiator 314 delayed by one symbol time period. Threshold T3 can be determined based on the application requirements, particularly the operational signal-to-noise ratio. In one implementation, the threshold T3 is set to of the maximum value which can be achieved in the sine wave represented by δδθn. For example, with reference to FIG. 11, T3 may be set to the value MAX.

An example implementation of pattern detection logic 316 is illustrated in FIG. 8. As illustrated, the samples δδθn are input in parallel to summer 802, and the series combination of L storage elements 800(0), 800(1), . . . , 800(L-1). The storage elements are configured such that successive samples of δδθn pass through each of the storage elements in succession, and the output of the Lth shift register is δδθn-L, a previous version of δδθn. As discussed previously, the number L is such that δδθn-L and δδθn are spaced by one symbol time period T.

The values δδθn and δδθn-L are input to summer 802. The output of summer 802 is input to block 804, which determines the absolute value of the output of summer 802. The output of block 804 is input to comparator 806, which compares this absolute value with a threshold T3. If this threshold is not exceeded, comparator 806 signals this condition on its output signal.

Meanwhile, the value δδθn-L output from storage element 800(L-1) is input to block 810, and the value δδθn is input to block 814. Block 810 determines the sign of δδθn-L, and signals the same on its output. Similarly, block 814 determines the sign of δδθn, and signals the same on its output. The outputs of blocks 810 and 814 are input to XOR block 812, which performs an exclusive OR function to determine if the signs of the two values δδθn and δδθn-L are the same or different. If the two are different, XOR block 812 signals this condition on its output signal.

The outputs of comparator 806 and XOR block 812 are input to AND block 808. AND block 808 determines whether the two conditions are satisfied whereby the absolute value of the sum of δδθn and δδθn-L is less than the predetermined threshold T3, and the signs of δδθn and δδθn-L are different. If both conditions are satisfied, AND block 808 signals this situations on its output signal dn.

Turning back to the burst detector of FIG. 3, moving window accumulator 318 sums the output dn of the pattern detection logic 316 over N samples, where the number N is the number of samples represented by the packet preamble. The summed output from the moving window accumulator 318 is designated as Sn in the figure. In current Bluetooth implementations, the number N is 16, representing 4 samples/symbol over 4 symbols, consistent with a 4 MHz sampling rate.

Moving window accumulator 318 may be implemented as shown in FIG. 9. As shown, the signal dn from pattern detection logic 316 is input to a series combination of N storage elements 900(0), 900(1), . . . , 900(N-1). The N storage elements are configured so that successive values of dn pass through each of the delay elements in succession. The outputs of the storage elements are each input to summer 902, which outputs as Sn the sum of the contents of the N storage elements.

In the burst detector of FIG. 3, peak detector 320 detects whether the output Sn of the moving window accumulator 318 has reached a peak and exceeds a predetermined threshold, signifying that a predetermined pattern of symbols has been detected in a packet preamble. If so, the peak detector 320 signals the occurrence of this condition on the output signal Pn.

In one implementation, the determination of Pn by the peak detector 320 can be represented by the following pseudo-code:
If (S n-1 >=S n and S n-1 >=S n-2 and S n-1 >T 4) then
Pn=high
Else
Pn low
where T4 is a predetermined threshold whose specific value depends on application requirements such as operational SNR, false alarm rate and misdetection rate. In one implementation, the threshold T4 is set to 12–14.

FIG. 10 illustrates a block diagram of this implementation of peak detector 320. As illustrated, the signal Sn is input to the series combination of storage elements 1000 and 1002. These storage elements are configured such that successive values of Sn pass successively through each of the storage elements. Consequently, Sn-1 is an output from storage element 1000, and Sn-2 is output from storage element 1002. The signals Sn, Sn-1, and Sn-2 are each input to decision module 1004 which is configured to determine the output Pn in accordance with the above pseudo-code.

Turning back to the burst detector of FIG. 3, burst detection logic 332 receives as inputs the output signal of the peak detector 320, and the output signal of the power change detection logic 330. If the output of the moving window accumulator 318 has peaked, as detected by the peak detector 320, and if both short-term and long-term power changes in the incoming signal have occurred as detected by the power change detection logic 330, burst detection logic 332 signals the occurrence of a burst. The occurrence of a burst (a) indicates an accurate burst starting point for timing/frequency acquisition in the packet preamble; and (b) activates block 310 to begin data demodulation. If one implementation, burst detection logic 332 simply signals the occurrence of a burst if both Dn and Pn are high. Otherwise the burst is not detected.

The acquisition and demodulation block 310 receives as an input the output δθn of the differential phase detector 308 after passage through the delay compensation buffer 310. The delay compensation buffer 310 is configured to compensate for the total delay inherent in the burst detection circuitry between the time that a packet is received and a burst is detected. Upon detection of a burst, the acquisition and demodulation block 310 is configured to acquire timing and frequency, and begin demodulating the differential phase input δδθn from the differential phase detector 306. That permits the underlying data 312 in the bodies of the packets (the payload filed 206 in FIG. 2) to be recovered.

FIG. 12 is a flowchart illustrating an embodiment of a method of operation of a burst detector in accordance with the invention. As shown, after initiation of the process, steps 1202, 1206, and 1210 are performed in parallel.

In step 1202, short-term power changes in the received signal are monitored. Step 1202 is followed by decision block 1204, where any short-term power change detected in step 1202 is compared with a predetermined threshold. If the threshold is not exceeded, a jump is made back to step 1202 for another iteration. If the threshold is exceeded, a jump is made to step 1214.

In step 1206, long-term power changes in the received signal are monitored. Step 1206 is followed by decision block 1208, where any long-term power change detected in step 1206 is compared with a predetermined threshold. If the threshold is not exceeded, a jump is made back to step 1206 for another iteration. If the threshold is exceeded, a jump is made to step 1214.

In step 1210, the pattern of symbols in the received signal is monitored. Step 1210 is followed by decision block 1212, where any pattern of symbols detected in step 1210 is compared with a predetermined pattern. If the predetermined pattern is not detected, a jump is made back to step 1210 for another iteration. If the predetermined pattern is detected, a jump is made to step 1214.

In step 1214, a burst detection is signaled if all three of the conditions represented by decision blocks 1204, 1208, and 1212 is satisfied, namely (a) that a short-term power change of sufficient magnitude has been detected; (b) that a long-term power change of sufficient magnitude has been detected; and (c) that a predetermined pattern of symbols has been detected.

Upon detection of a burst, depending on the application, various steps such as frequency and/or timing acquisition, or demodulation, may be allowed to occur. The process may then iterate in order to detect additional bursts.

The foregoing method may be embodied in the form of computer readable media, memory, or circuitry, which tangibly embodies the foregoing method in the form of instructions or some other form.

In addition, it should be appreciated by one of ordinary skill in the art that, in lieu of determining if the power changes exceed predetermined thresholds, steps 1204 and 1208 may involve determining if the power changes equal or exceed predetermined thresholds, or some other method of determining if the power changes are of a sufficient magnitude to distinguish over noise.

It should further be appreciated that embodiments are possible in which separate short-term and long-term power changes are not monitored, where only a single power change is monitored, or where two or more power changes are monitored which differ by other than time horizon.

From the foregoing, it should be appreciated that a burst detection system has been described and illustrated which is capable of detecting bursts represented by packets with relatively short preambles, although the application of the system is not so limited.

Another advantage is accurate detection of bursts through the joint operation of the one or more power change detectors and the one or more pattern detectors, whereby the one or more power change detectors reduce the chance of a false negative condition, i.e., the failure to detect a burst at all, while the one or more pattern detectors reduce or eliminate the chance of a false positive condition, i.e., the false detection of a burst.

Further advantages will be apparent to those of ordinary skill in the art from a reading of this disclosure.

While various embodiments of the invention have been described, it will be apparent to those of ordinary skill in the art that many more embodiments and implementations are possible that are within the scope of this invention. Accordingly, the invention is not to be restricted except in light of the attached claims and their equivalents.

Patent Citations
Cited PatentFiling datePublication dateApplicantTitle
US5272446 *Aug 17, 1992Dec 21, 1993ComsatDigitally implemented fast frequency estimator/demodulator for low bit rate maritime and mobile data communications without the use of an acquisition preamble
US5590410 *Oct 26, 1994Dec 31, 1996American Wireless CorporationSystem and method for frequency based acquisition acknowledgment between transmitter and receiver
US5621764 *Jun 21, 1995Apr 15, 1997Nec CorporationSoft decision signal outputting receiver
US5621766 *Oct 31, 1994Apr 15, 1997Motorola, Inc.Method and apparatus for burst detecting
US5802076 *May 24, 1996Sep 1, 1998National Semiconductor CorporationAudio error mitigation technique for a TDMA communication system
US5995816 *Dec 22, 1997Nov 30, 1999Nec CorporationFrequency hopping communications receivers
US6272186 *Dec 22, 1998Aug 7, 2001Nortel Networks LimitedNormal burst acquisition system for use in a cellular communications network
US6370210 *May 21, 1999Apr 9, 2002Nec CorporationCircuitry for generating a gain control signal applied to an AGC amplifier and method thereof
US6603746 *Jun 18, 1999Aug 5, 2003Nortel Networks LimitedMethod and apparatus for controlling transmitted power in a wireless communications system
US6690681 *Mar 21, 2000Feb 10, 2004Airbiquity Inc.In-band signaling for data communications over digital wireless telecommunications network
Referenced by
Citing PatentFiling datePublication dateApplicantTitle
US7187724 *Nov 1, 2002Mar 6, 2007Advanced Micro Devices, Inc.Average tracking mechanism in data communications receivers
US7218628 *Jul 29, 2002May 15, 2007Mediatek IncorporationMethod and device for detecting preamble of wireless data frame
US7751518 *Feb 1, 2006Jul 6, 2010Sunplus Technology Co., Ltd.System and method for executing preamble detection, symbol timing recovery, and frequency offset estimation
US7769352 *Sep 25, 2006Aug 3, 2010Kabushiki Kaisha ToshibaReceiver and wireless communication apparatus
US8254482 *Aug 28, 2012Samsung Electronics Co., Ltd.Perturbed decoder, perturbed decoding method and apparatus in communication system using the same
US8335283 *Nov 11, 2008Dec 18, 2012Qualcomm Atheros, Inc.Weak signal detection in wireless communication systems
US20030147374 *Jul 29, 2002Aug 7, 2003Mao-Ching ChiuMethod and device for detecting preamble of wireless data frame
US20040001562 *Nov 1, 2002Jan 1, 2004Menno MennengaAverage tracking mechanism in data communications receivers
US20070071136 *Feb 1, 2006Mar 29, 2007Sunplus Technology Co., Ltd.System and method for executing preamble detection, symbol timing recovery, and frequency offset estimation
US20080254756 *Sep 25, 2006Oct 16, 2008Kabushiki Kaisha ToshibaReceiver and Wireless Communication Apparatus
US20090285342 *May 13, 2008Nov 19, 2009Suh Jung HoonPerturbed decoder, perturbed decoding method and apparatus in communication system using the same
Classifications
U.S. Classification375/342, 375/341, 375/317
International ClassificationH04L25/06, H04L7/04, H04B17/00
Cooperative ClassificationH04L25/062, H04L7/042, H04L2007/047, H04B17/318
European ClassificationH04L7/04B1, H04L25/06A1, H04B17/00B1R
Legal Events
DateCodeEventDescription
Feb 28, 2001ASAssignment
Owner name: CONEXANT SYSTEMS, INC., CALIFORNIA
Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNORS:YANG, GANNING;WALLEY, JOHN;REEL/FRAME:011580/0336
Effective date: 20010205
Sep 12, 2002ASAssignment
Owner name: ALPHA INDUSTRIES, INC., MASSACHUSETTS
Free format text: MERGER;ASSIGNOR:WASHINGTON SUB, INC.;REEL/FRAME:013288/0249
Effective date: 20020625
Owner name: SKYWORKS SOLUTIONS, INC., CALIFORNIA
Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNOR:ALPHA INDUSTRIES, INC.;REEL/FRAME:013288/0241
Effective date: 20020625
Owner name: WASHINGTON SUB, INC., CALIFORNIA
Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNOR:CONEXANT SYSTEMS, INC.;REEL/FRAME:013288/0197
Effective date: 20020625
Sep 16, 2002ASAssignment
Owner name: CONEXANT SYSTEMS, INC., CALIFORNIA
Free format text: SECURITY INTEREST;ASSIGNOR:ALPHA INDUSTRIES, INC.;REEL/FRAME:013240/0860
Effective date: 20020625
Jan 13, 2003ASAssignment
Owner name: SKYWORKS SOLUTIONS, INC., CALIFORNIA
Free format text: CORRECTIVE MERGER ON REEL 013288 FRAME 0241 TO DELETE S/N 09/811,082 AND 09/793,744;ASSIGNOR:ALPHA INDUSTRIES, INC.;REEL/FRAME:013645/0527
Effective date: 20020626
Oct 6, 2003ASAssignment
Owner name: ALPHA INDUSTRIES, INC., MASSACHUSETTS
Free format text: RELEASE AND RECONVEYANCE/SECURITY INTEREST;ASSIGNOR:CONEXANT SYSTEMS, INC.;REEL/FRAME:014580/0880
Effective date: 20030307
Jul 31, 2009FPAYFee payment
Year of fee payment: 4
Jul 31, 2013FPAYFee payment
Year of fee payment: 8